I’m trying to better understand the behavior of enabling both stagger and reapply under execution behavior. We’re attempting to perform an action with the following behavior:
- Runs between 4AM-6AM client local time
- Reapply unlimited number of times, waiting 1 day between reapplications
- Stagger over 90 minutes
Each target system is expected to become relevant again within 3 days. However, the problem is that the stagger function isn’t working as expected. Whenever the action reapplies, most devices begin right at the beginning of the constraint window (4AM) rather than the randomly distributed start delay (sometime between 4AM-5:30AM).
It seems like the stagger only applies to the first application and subsequent reapplications no longer delay the start time. Or perhaps I’m misunderstanding, but either way is there any recourse? I could probably find another way to stagger the action outside of the execution behavior if needed.